Chiral Nanozymes-Gold Nanoparticle-Based Transphosphorylation Catalysts Capable of Enantiomeric Discrimination

Article abstract of DOI:10.1002/chem.201600853

Enantioselectivity in RNA cleavage by a synthetic metalloenzyme has been demonstrated for the first time. Thiols containing chiral Zn^II-binding head groups have been self-assembled on the surface of gold nanoparticles. This results in the spontaneous formation of chiral bimetallic catalytic sites that display different activities (k_cat) towards the enantiomers of an RNA model substrate. Substrate selectivity is observed when the nanozyme is applied to the cleavage of the dinucleotides UpU, GpG, ApA, and CpC, and remarkable differences in reactivity are observed for the cleavage of the enantiomerically pure dinucleotide UpU.

Full text of DOI:10.1002/chem.201600853

DOI: 10.1002/chem.201600853
Communication
&
Nanozymes
Chiral Nanozymes–Gold Nanoparticle-Based Transphosphorylation
Catalysts Capable of Enantiomeric Discrimination
Jack L.-Y. Chen, Cristian Pezzato, Paolo Scrimin, and Leonard J. Prins*[a]
Chem. Eur. J. 2016, 22, 7028 – 7032
7028
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
enantioselectivity in the cleavage of UpU. The ability to display
enantioselective catalysis brings these synthetic nanosystems
another step closer to mimicking enzymes.
Abstract: Enantioselectivity in RNA cleavage by a synthetic
metalloenzyme has been demonstrated for the first time.
Thiols containing chiral ZnII-binding head groups have
been self-assembled on the surface of gold nanoparticles.
This results in the spontaneous formation of chiral bimet-
allic catalytic sites that display different activities (kcat) to-
wards the enantiomers of an RNA model substrate. Sub-
strate selectivity is observed when the nanozyme is ap-
plied to the cleavage of the dinucleotides UpU, GpG, ApA,
and CpC, and remarkable differences in reactivity are ob-
served for the cleavage of the enantiomerically pure dinu-
cleotide UpU.
We have prepared two batches of NPs (Figure 1), each
coated with a self-assembled monolayer of enantiomerically
pure thiols where the chirality of the head group is derived
from either l or d serine.[8] To ensure nanoparticle uniformity,
Chemists have long been fascinated with the complexity and
efficiency of Nature’s enzymes. This has led to the develop-
ment of numerous synthetic systems that try to emulate the
properties of this natural machinery.[1] Recently, gold nanoparti-
cles (AuNPs) passivated by thiols which terminate with metal
complexes[2] have been shown to possess many features of
metalloenzymes.[3] These include the cooperativity between
neighboring metal ions and Michaelis–Menten kinetics in the
cleavage of phosphodiester bonds.[2b] These AuNPs have thus
emerged as an attractive platform for the development of syn-
thetic enzymes[4] and, in particular, as a model[5] for nucleases.
However, despite the large body of work dedicated to synthet-
ic metalloenzymes for RNA cleavage,[2,5,6] there are to the best
of our knowledge no studies aimed at enantiodifferentiation.
This is a remarkable fact considering that enantioselectivity is
one of the hallmarks of enzyme catalysis.
Figure 1. Schematic representation of the self-assembly of thiols containing
chiral head groups on the surface of dioctylamine-passivated gold nanopar-
ticles to form (+)-1 and (À)-1 NPs, which are active for the catalysis of
CF3HPNP. CF3HPNP=2-hydroxypropyl p-nitro-m-trifluoromethylphenyl phos-
phate. Only one hemisphere of the (+)-1 and (À)-1 NPs is shown.
Our aim was to employ AuNPs as a platform on which to
self-assemble chiral precursors and generate an active catalyst
capable of enantiodiscrimination. This goal is fundamentally
different to the immobilization of known and established chiral
catalysts onto gold nanoparticles (or other solid supports).[7] In
our proposed system, the individual thiols only represent half
the catalytic site, and its self-assembly on the surface of the
gold nanoparticle is essential for formation of the active cata-
lyst. In fact, it has been shown previously that the individual
metal complexes are poorly active for the transphosphoryla-
tion of a RNA model substrate when free in solution,[2c] giving
further importance to the self-assembly process. Herein, we
report that chiral metal-binding thiols are indeed able to self-
assemble on the surface of AuNPs to form active catalysts and
exhibit different reactivity for the enantiomers of a pure RNA
model substrate. Subsequently, by taking advantage of the
multivalent nature of the nanozyme, additional binding inter-
actions with dinucleotides were exploited to induce substrate
selectivity in the cleavage of UpU, GpG, ApA, and CpC and
the two batches of NPs (+)-1 and (À)-1 were synthesized from
a single batch of dioctylamine-coated NPs prepared using
a modified version of the procedure by Peng et al.[9] This in-
volved initial stabilization of the gold nanoparticles by
a weakly coordinating ligand, which allowed for uniform size
distribution, prior to final stabilization by the addition of thiols.
The resulting well-distributed sub-2 nm gold NPs were split
into two, before introduction of the corresponding enantio-
meric thiols to each batch. The final NPs (+)-1 and (À)-1 were
analyzed by 1H NMR, UV spectroscopy, transmission electron
microscopy (TEM), and dynamic light scattering (DLS) and
found to be identical (see the Supporting Information). Ther-
mogravimetric analysis also confirmed that the loading of
thiols onto (+)-1 and (À)-1 NPs were the same. Critically, circu-
lar dichroism (CD) measurements displayed equal and opposite
signals at 204 nm for NPs (+)-1 and (À)-1, reflecting the pres-
ence of thiols of opposite chirality on the NP surface.
The thiols employed in the above-mentioned self-assembly
terminate with triazacyclononane (TACN) metal-binding units.
The resulting complexes with ZnII are the crucial elements for
cleaving the phosphodiester moiety of an RNase substrate.
Previous studies have shown that at least two neighboring
metal ions work cooperatively (see Figure 1).[5f] At this point,
we needed to synthesize enantiomerically pure RNA model
[a] Dr. J. L.-Y. Chen, Dr. C. Pezzato, Prof. Dr. P. Scrimin, Prof. Dr. L. J. Prins
Department of Chemical Sciences, University of Padova
Via Marzolo 1, 35131 Padova (Italy)
E-mail: leonard.prins@unipd.it
Supporting information for this article and ORCID for the corresponding
author are available on the WWW under
http://dx.doi.org/10.1002/chem.201600853.
Chem. Eur. J. 2016, 22, 7028 – 7032
www.chemeurj.org
7029
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
substrates to verify whether the NPs (+)-1 and (À)-1 were ca-
pable of enantioselective catalysis.
was followed using the opposite enantiomer of alcohol 3 for
the synthesis of the enantiomeric (À)-CF3HPNP, (À)-2.
We chose to pursue the synthesis of 2-hydroxypropyl p-
nitro-m-trifluoromethylphenyl phosphate (+)-2 (CF3HPNP;
Scheme 1), a trifluoromethyl-substituted analogue of 2-hy-
droxypropyl p-nitrophenyl phosphate (HPNP). This more reac-
tive substrate gave greater flexibility when probing different
reaction conditions (such as changes in pH) to gauge any dif-
ferences in the catalytic activity of the NPs (+)-1 and (À)-1.
The required chiral substrate was synthesized beginning with
the condensation of 3-trifluoromethyl-4-nitrophenol and phos-
phorus oxychloride (Scheme 1),[10] followed by further conden-
sation with chiral alcohol 3,[6d] which provided the optically
pure chiral center. Desilylation of the resulting triethylamine
salt and cation exchange then provided the desired optically
pure (+)-CF3HPNP (+)-2 as a Na+ salt. This same procedure
Initial studies in aqueous buffer, using the nanoparticle cata-
lysts ([TACN·ZnII]=5 mm) and the (+)-CF3HPNP substrate (2 mm)
immediately garnered interesting results. A plot of the absorb-
ance at 405 nm, corresponding to the presence of product, re-
vealed a clear difference in the hydrolysis rate of (+)-CF3HPNP
in the presence of either (+)-1 or (À)-1 NPs (Figure 2a). Fitting
of the curves to pseudo-first-order kinetics revealed the kobs
value for the hydrolysis in the presence of (À)-1 NP to be
1.1Æ0.1”10À4 sÀ1, almost half the kobs in the presence of (+)-1
NP (2.0Æ0.1”10À4 sÀ1). Control experiments using the enantio-
meric (À)-CF3HPNP with NPs (+)-1 and (À)-1, revealed an in-
version in the curves obtained from the above experiment,
with a kobs of 2.1Æ0.1”10À4 sÀ1 observed with (À)-1 NP and
1.1Æ0.1”10À4 sÀ1 observed with (+)-1 NPs. This important
control using the opposite enantiomer of the original substrate
left no doubt that the observed difference in reaction rate was
indeed due to a difference in the chiral environment on the
surface of the gold nanoparticles.
A surprising insight into the origin of this difference was ob-
tained from a study of the catalytic activity of (+)-1 and (À)-
1 NPs at varying concentrations of the (+)-CF3HPNP substrate
([TACN·ZnII]=10 mm, pH 6.5). A plot of the initial reaction rates
with respect to (+)-CF3HPNP concentration revealed signifi-
cantly different profiles between (+)-1 and (À)-1 NPs (Fig-
ure 2b). Fitting of the saturation profiles to the Michaelis–
Menten equation yielded a kcat value of 2.5Æ0.1”10À3 sÀ1 for
(À)-1 NPs, whereas for (+)-1 NPs a 38% higher value was ob-
Scheme 1. Synthesis of chiral substrate (+)-CF3HPNP.[6d,10]
Figure 2. a) Catalysis of (+)- or (À)-CF3HPNP (2 mm) by (+)-1 or (À)-1 NPs ([TACN]=5 mm), as a percentage of conversion over time; HEPES pH 7.0 buffer
(10 mm), Zn(NO3)2 (5 mm), 408C. b) Plot of initial reaction rates versus [CF3HPNP], showing fit with Michaelis–Menten kinetics; MES pH 6.5 buffer (10 mm),
(+)-1 or (À)-1 NPs (10 mm), Zn(NO3)2 (10 mm), (+)- or (À)-CF3HPNP (variable), 408C. c) Inhibition studies using dimethyl phosphate as inhibitor; MES pH 6.5
buffer (10 mm), (+)-1 or (À)-1 NPs (10 mm), Zn(NO3)2 (10 mm), dimethylphosphate (variable), 408C. d) Summary of Michaelis–Menten data at different pH
values. See Supporting Information for further details.
Chem. Eur. J. 2016, 22, 7028 – 7032
www.chemeurj.org
7030
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
tained (kcat =3.5Æ0.1”10À3 sÀ1). On the contrary, similar KM
(which is as a measure of binding affinity) values of 0.24Æ
0.02 mm and 0.27Æ0.02 mm were obtained for (À)-1 and
(+)-1 NPs, respectively. This is remarkable, considering that
a different affinity of the chiral substrate for the enantiomeric
NPs would be the most obvious explanation for the observed
difference. However, considering that the KM values are un-
changed—this is clearly not the case. This was further con-
firmed by an inhibition experiment in which the catalytic activ-
ity was measured using substrate (+)-CF3HPNP in the presence
of increasing amounts of an achiral inhibitor—dimethylphos-
phate (Figure 2c). It can be seen that the extent of inhibition is
the same for both (+)-1 and (À)-1 NPs, which confirms that
(+)-CF3HPNP has the same affinity for both NPs. Remarkably,
the difference in catalytic activity originates from the difference
in kcat between (+)-1 and (À)-1 NPs (38% difference at pH 6.5).
In enzymatic catalysis, kcat is a first-order rate constant that
refers to the efficiency of the enzyme in catalyzing the chemi-
cal transformation and is often referred to as the turnover
number. This implies that the difference in catalytic activity can
be attributed to energy differences in the respective transition
states that are formed when (+)-CF3HPNP reacts on the surface
of (+)- and (À)-1 NPs. This is an exciting observation as it
demonstrates that the observed selectivity originates from dif-
ferences in the transition state, rather than the ground state.
Again, crucial control experiments were performed with the
enantiomeric substrate ((À)-CF3HPNP), which demonstrated
a reversal in the hydrolysis kinetics (see Figure 2b).
The effect of pH on the kinetics of the reaction was also in-
vestigated. On increasing the pH from 6.5 to 8.0, an increase in
catalytic activity (kcat) is observed, but the difference in kcat
from using (+)-1 or (À)-1 NPs decreases from 38% at pH 6.5
down to 21% at pH 8.0 (Figure 2d). The KM also shows a gener-
al increase going from pH 7.0 to 8.0, due likely to increased
competition from the negatively charged form of the buffer
used (HEPES or MES) binding at the surface of the NPs. In all
cases, at the different pH values no significant differences in KM
were observed for the (+)-1 and (À)-1 NPs.
Figure 3. a) Transphosphorylation of dinucleotides. b) Tansphosphorylation
of NpN with (+)-1 NPs, showing fractional conversion over time. c) Cleavage
of UpU with (+)-1 and (À)-1 NPs in the presence of CuII; in HEPES pH 8.0,
room temperature. d) Summary of rate constants with the different condi-
tions reported. See Supporting Information for further details.
tremely slow, ranging from a rate constant of 9.8”10À9 sÀ1 for
UpU[6i] (at pH 7) to 1.7”10À9 sÀ1 for ApA.[14]
Upon demonstrating with model substrates that enantiose-
lective catalysis is possible with our nanoparticle system, the
next goal was to examine ways to increase the enantioselectiv-
ity achieved. We reasoned that it was crucial to exploit the
multivalency of our system for the formation of additional in-
teractions between the catalyst and the substrate. This would
allow precise binding conformations to be formed, giving im-
petus for enantioselectivity to occur. RNA dinucleotides (3’,5’-
NpN, N=U, C, G, A; Figure 3a) emerged as ideal substrates to
study,[11] as multiple simultaneous interactions are believed to
occur between the metal ions and both the base and phos-
phate moieties.[12] Aza-crown-ZnII complexes similar to our
TACN·ZnII catalysts have been shown to bind specific base moi-
eties, with a special preference for thymine and uracil.[13] The
cleavage of these dinucleotides can be followed by HPLC,
monitoring the disappearance of the substrates as well as the
formation of cyclic ribonucleoside monophosphate (2’,3’-
cNMP) and the corresponding nucleoside. The uncatalyzed
cleavage of these compounds (UpU, CpC, GpG, and ApA) is ex-
The cleavage of the dinucleotides with (À)-1 NPs ([TACN·Z-
nII]=50 mm) was investigated at pH 8.0. Importantly, the cata-
lysts demonstrated significant substrate specificity in the cleav-
age of the dinucleotides, promoting the cleavage of UpU with
a second-order rate constant of 2.50Æ0.07”10À2 sÀ1 mÀ1 (Fig-
ure 3b). This rate was 3.6 ” faster than for GpG (6.7Æ0.5”
10À3 sÀ1 mÀ1), which in turn reacted 10 ” faster than CpC (9.4Æ
0.2”10À4 sÀ1 mÀ1) and ApA (7.0Æ0.2”10À4 sÀ1 mÀ1). The ob-
served selectivity for UpU is in line with the previously report-
ed recognition between related ZnII complexes and uracil (see
above).[12] Due to its observed higher activity, further investiga-
tions looking at the difference in catalytic activity between
(+)-1 and (À)-1 NPs were thus performed using UpU as the
substrate. Initial trials looking at the difference in catalytic ac-
tivity between (+)-1 and (À)-1 NPs with UpU showed no sig-
nificant difference when ZnII was employed as the metal ion
(Figure 3d). However, a remarkable result was observed when
the divalent metal used in the catalysis was switched from ZnII
Chem. Eur. J. 2016, 22, 7028 – 7032
www.chemeurj.org
7031
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
to CuII. In this case we observed that at pH 7.5, (À)-1 NPs
(13.2Æ0.2”10À4 sÀ1) were able to catalyze the transphosphory-
[1] a) M. Raynal, P. Ballester, A. Vidal-Ferran, P. W. N. M. van Leeuwen, Chem.
Soc. Rev. 2014, 43, 1734; b) Y. H. Lin, J. S. Ren, X. G. Qu, Acc. Chem. Res.
2014, 47, 1097; c) H. Wei, E. K. Wang, Chem. Soc. Rev. 2013, 42, 6060;
d) R. Bhandari, R. Coppage, M. R. Knecht, Catal. Sci. Technol. 2012, 2,
256; e) Supramolecular Catalysis (Ed.: P. W. N. M. van Leeuwen), Wiley-
VCH, Weinheim, 2008; f) D. M. Vriezema, M. C. Aragones, J. A. A. W. Ele-
mans, J. J. L. M. Cornelissen, A. E. Rowan, R. J. M. Nolte, Chem. Rev. 2005,
105, 1445.
[2] a) M. Diez-Castellnou, F. Mancin, P. Scrimin, J. Am. Chem. Soc. 2014, 136,
1158; b) G. Zaupa, C. Mora, R. Bonomi, L. J. Prins, P. Scrimin, Chem. Eur.
J. 2011, 17, 4879; c) F. Manea, F. B. Houillon, L. Pasquato, P. Scrimin,
Angew. Chem. Int. Ed. 2004, 43, 6165; Angew. Chem. 2004, 116, 6291.
[3] N. Sträter, W. N. Lipscomb, T. Klabunde, B. Krebs, Angew. Chem. Int. Ed.
Engl. 1996, 35, 2024; Angew. Chem. 1996, 108, 2158.
lation of UpU 3.4
”
faster than (+)-1 NPs (3.9Æ0.1”
10À4 sÀ1 mÀ1; Figure 3d). This difference was further increased
at pH 8, with a fourfold difference in rate observed between
(À)-1 NPs (17.4Æ0.1”10À4 sÀ1) and (+)-1 NPs (4.3Æ0.1”
10À4 sÀ1 mÀ1; Figure 3c). The rate observed with (+)-1 NPs is es-
sentially the same as the background rate (control). Thus (+)-1
NPs are effectively not catalyzing the hydrolysis of UpU at all.
These results represent the first example for the enantioselec-
tive cleavage of dinucleotides using synthetic metalloenzymes.
This has been achieved using the self-assembly of precursors
which are by themselves catalytically inactive. The demonstra-
tion of enantioselectivity in these biomimetic metallozymes
comes one step closer to mimicking the complexity and prop-
erties of Nature’s own enzymes. By addressing this issue of
chirality, we illustrate the opportunities of using self-assembly
for the generation of novel chiral synthetic catalysts and the
vast potential of this class of nanozymes for further develop-
ment.
[4] a) D. Zaramella, P. Scrimin, L. J. Prins, J. Am. Chem. Soc. 2012, 134, 8396;
b) P. Pengo, S. Polizzi, L. Pasquato, P. Scrimin, J. Am. Chem. Soc. 2005,
127, 1616.
[5] a) E. Y. Tirel, Z. Bellamy, H. Adams, V. Lebrun, F. Duarte, N. H. Williams,
Angew. Chem. Int. Ed. 2014, 53, 8246; Angew. Chem. 2014, 126, 8385;
b) F. Mancin, P. Scrimin, P. Tecilla, Chem. Commun. 2012, 48, 5545; c) C. T.
Liu, A. A. Neverov, R. S. Brown, J. Am. Chem. Soc. 2008, 130, 16711;
d) C. T. Liu, A. A. Neverov, R. S. Brown, J. Am. Chem. Soc. 2008, 130,
13870; e) A. A. Neverov, Z. L. Lu, C. I. Maxwell, M. F. Mohamed, C. J.
White, J. S. W. Tsang, R. S. Brown, J. Am. Chem. Soc. 2006, 128, 16398;
f) N. H. Williams, B. Takasaki, M. Wall, J. Chin, Acc. Chem. Res. 1999, 32,
485; g) R. Breslow, J. B. Doherty, G. Guillot, C. Lipsey, J. Am. Chem. Soc.
1978, 100, 3227; h) N. H. Williams, A. M. Lebuis, J. Chin, J. Am. Chem.
Soc. 1999, 121, 3341.
[6] a) R. Salvio, S. Volpi, R. Cacciapaglia, A. Casnati, L. Mandolini, F. Sansone,
J. Org. Chem. 2015, 80, 5887; b) H. Lçnnberg, Org. Biomol. Chem. 2011,
9, 1687; c) Y. Aiba, J. Sumaoka, M. Komiyama, Chem. Soc. Rev. 2011, 40,
5657; d) T. Humphry, S. Iyer, O. Iranzo, J. R. Morrow, J. P. Richard, P.
Paneth, A. C. Hengge, J. Am. Chem. Soc. 2008, 130, 17858; e) J. R.
Morrow, T. L. Amyes, J. P. Richard, Acc. Chem. Res. 2008, 41, 539; f) T.
Niittymäki, H. Lonnberg, Org. Biomol. Chem. 2006, 4, 15; g) J. R. Morrow,
O. Iranzo, Curr. Opin. Chem. Biol. 2004, 8, 192; h) P. Molenveld, J. F. J.
Engbersen, D. N. Reinhoudt, Angew. Chem. Int. Ed. 1999, 38, 3189;
Angew. Chem. 1999, 111, 3387; i) W. H. Chapman, R. Breslow, J. Am.
Chem. Soc. 1995, 117, 5462; j) J. Chin, Curr. Opin. Chem. Biol. 1997, 1,
514.
[7] G. Pieters, L. J. Prins, New J. Chem. 2012, 36, 1931; a notable example is
the hydrolytic kinetic resolution of epoxides by Jacobsen et al. where
catalytic activity can be increased by immobilization of catalysts onto
NPs: T. Belser, E. N. Jacobsen, Adv. Synth. Catal. 2008, 350, 967.
[8] P. Rossi, F. Felluga, P. Scrimin, Tetrahedron Lett. 1998, 39, 7159.
[9] a) F. Manea, C. Bindoli, S. Polizzi, L. Lay, P. Scrimin, Langmuir 2008, 24,
4120; b) N. R. Jana, X. G. Peng, J. Am. Chem. Soc. 2003, 125, 14280.
[10] J. P. Horwitz, J. V. Freisler, J. Med. Chem. 1970, 13, 1024.
[11] a) H. Linjalahti, G. Q. Feng, J. C. Mareque-Rivas, S. Mikkola, N. H. Williams,
J. Am. Chem. Soc. 2008, 130, 4232; b) D. M. Perreault, E. V. Anslyn,
Angew. Chem. Int. Ed. Engl. 1997, 36, 432; Angew. Chem. 1997, 109, 470;
c) M. Oivanen, S. Kuusela, H. Lonnberg, Chem. Rev. 1998, 98, 961.
[12] a) Q. Wang, H. Lonnberg, J. Am. Chem. Soc. 2006, 128, 10716; b) Q.
Wang, E. Leino, A. Jancso, I. Szilagyi, T. Gajda, E. Hietamaki, H. Lonnberg,
ChemBioChem 2008, 9, 1739.
In conclusion, we have shown that chirality is an important
design criterion when it comes to assembling nanozymes.
While this has so far been overlooked for this class of struc-
tures, we show that chiral precursors indeed are able to self-or-
ganize on the surface of nanoparticles and form chiral catalyti-
cally active pockets. While we are still far away from the per-
formance of an enzyme, we have made a significant step for-
ward by showing for the first time that this class of enzymes is
capable of enantiodiscrimination both on model and natural
compounds. We show also that the multivalent nature of this
system can be expoited to provide enhanced levels of enantio-
selectivity. This work presents a shift away from the incorpora-
tion of an inherently chiral catalyst on the surface of NPs to
the self-assembly of chiral catalytic pockets from units that by
themselves do not induce enantioselectivity. These systems
have the advantages of modularity and greater flexibility in
design. As such, this is a major advancement for the develop-
ment of innovative catalytic systems for enantioselective catal-
ysis.
Acknowledgements
Financial support from the ERC (StG-239898), University of
Padova (CPDA138148) and COST (CM1304) is acknowledged.
[13] a) S. Aoki, E. Kimura, Chem. Rev. 2004, 104, 769; b) M. Shionoya, T. Ikeda,
E. Kimura, M. Shiro, J. Am. Chem. Soc. 1994, 116, 3848.
[14] M. Komiyama, K. Yoshinari, J. Org. Chem. 1997, 62, 2155.
Keywords: chirality · enzyme mimics · nanozymes · self-
assembly · transphosphorylation
Received: February 23, 2016
Published online on April 8, 2016
Chem. Eur. J. 2016, 22, 7028 – 7032
www.chemeurj.org
7032
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim